Cobbled prospecting scintillation detector

[Jim Kovalchick]

Sorry not spectrometry or spectroscopy related, but there are some related concepts here.

Yesterday, I cobbled together a detector probe I intend to use for uranium ore prospecting, out of some parts I had laying around. My probe works, but I will admit that I was more than irresponsible in its design as I will mention more specifically directly. I am interested in opinions about fatal flaws.

I started with a small cylindrical BGO crystal I bought from someone at a science flea market for about 15 US dollars. It looks fine other than a couple very tiny chips on one edge. I wrapped all it but one end with some white Teflon tape.

Some years back someone gifted me a Hamamatsu R4125 pmt new in a box marked as a sample. The R4125 has a normal rating of -1500 V, a 0.27 mA divider current, and a max overall voltage of -1800. Its diameter is slightly larger than the crystal. I used some silicone grease to mate the pmt to the crystal. I covered the whole thing with 10 mil pvc tape for light blocking. I did not use any magnetic shielding.

I chose a basic positive bias voltage divider design because I was going to use the probe with a handheld ratemeter. The power supplies in normal handheld ratemeters aren't particularly strong and wanting to conserve battery power, I chose 6 Mohm resistors. I would have preferred at least 10 Mohm, but I didn't have enough of those around yesterday. I did not put zener diodes on the final stages even though I expect to see some high count rates. I used 10 nF caps on the final three stages.

I wrapped the assembly in a thin layer of foam and fitted it in a 1" pvc tube. I added a MHV bulkhead fitting on the base.

I attached the probed to a five foot MHV cable and an Eberline PRM 6 ratemeter. The PRM 6 has an adjustable HV but no adjustable discrimination or threshold. I already had the meter's HV adjusted to 900 V to start. The probe was detecting radiation immediately on power up. The background is about 1000 cpm. I found that for gamma, the probe registers counts 3 to 4 times what a two inch pancake geiger does. For example, on a relatively strong thorium source, the on-contact reading from a 2" pancake was around 40,000 cpm. For the home made BGO probe read 160,000 cpm. While reading this high count rate, I switched the meter over to indication of HV and noted that it was dragging down the ratemeter to 650 V from 900. Lower count rates do not have this effect.

Given the initial results, I have graded myself a C for the effort. I still intend to do some longer term testing to check battery life. I will probably also harden the probe for field work and mount it on a staff to get it closer to the ground. Thoughts anyone has on things I missed or failed at are welcome.

Thanks,

Jim K

[Sesselmann]

Jim,

Well done 👍, the probe seems perfectly adequate for the job, rugged probe for prospecting and if you find some hot rocks with it, does it really matter if there is a voltage drop? 10M resistors might have been better for a rate meter, but frankly it doesn't matter unless it's used to measure exposure.


Steven

[Jim Kovalchick]

Steven,
Thanks for the comments. While I haven't had the chance to test it in the field yet, the probe works well with both mild and spicy rocks in my collection. I've been running it now for at least a few hours and it hasn't expended the batteries yet, and that was one of my goals. It needs to last for a while in the field. I am going to leave it at my initial voltage setting for now, but I have quite a bit of bandwidth to play with if I want to adjust it to get optimum needle reaction that I can see in the field. Speaking of voltage adjustment, I believe that the old ratemeter I am using, the Eberline PRM 6, is an under-rated machine. It can count up to 500k cpm (I've tested it with this probe to 300k), and the voltage range can get to 1500 V. It has a MHV connector to handle that range. These machines don't have the well known reputation for cheap versatility that a Ludlum Model 3 has, but that only makes them a cheaper score on eBay. One down side is they are a little heavy. Using a strap is important. I am attaching a photo of my probe and the PRM 6 I paired it with.

By the way, I covered the outside of my probe with 10 mil tape for two reasons. The first is I expect to beat of the probe a little on rocks, and the other is to help with light. While I did cover my PMT with tape inside the housing, the white PVC the housing is made from is somewhat translucent.

The picture shows the background in my apartment. It's on the x10 scale and reading a little under a 1000 cpm. In the current configuration, I can pick up Am 241 gammas outside of a modern smoke detector when my 2" pancake cannot. The Geiger wins when the the smoke detector source is open because of the alphas. Similarly, it can pick up counts from Cs 137 in a fairly old spark gap tube that my pancake cannot see. A phosphate nodule containing a little uranium rings up 700 cpm on my pancake and yields around 3000 cpm with the new probe. My plan is to bring the new probe into the field with a Ludlum model 26 geiger slung on my shoulder with the face bagged . Together, the two make a good pair.

Cheers,

Jim K

[kotarak]

Hi Jim,
I'll suggest that you look into an Eberline ASP-1 meter - it has an adjustable click divider for the audio which makes the audio clicks actually useable with scintillators when prospecting.
You can set the ASP-1 to produce one click every 64, 128, 256, etc counts and since scintillators have high count rate undivided audio is basically a constant screech. If you go to a larger crystal and overall, a more sensitive detector you'll be getting even more counts and dividing the audio clicks is the only way to get a useable audio.

Here are my attempts at creating a bit more sophisticated prospecting instrument - http://blog.kotarak.net/2021/10/gamma-d ... -rock.html

I just completed a slightly updated version of the hardware - http://blog.kotarak.net/2023/07/gamma-d ... b-v40.html

[Jim Kovalchick]

Thanks. Your project looks great, and I agree with the overall premise that a click for each count doesn't work well with scintillators. I've never been a proponent of clicks anyway. For surveys, I much prefer a constant tone that changes pitch with count rate. A good example is the Ludlum 2220 (not 2221) scaler/ratemeter. I love it for surveying because the pitch change is very easy to pick up by ear with changing count rate. Further, the starting tone frequency resets when changing ranges. The only downside is the size of it. The 2220 is a beast, but I've thought about bringing it into the field instead of the PRM.

[kotarak]

The Gamma Dog instrument I posted links to is exactly that - it uses tone with variable pitch based on the count rate.
In addition there is an Automatic "Smart" Squelch System which monitors the background rate and sets the squelch threshold at background level so if a count rate above background is detected, the squelch will open and the variable pitch tone will sound. This way you dont have to constantly listen to variations of the rate if there is no reason.
The Smart squelch system will constantly adjust the threshold level and if you find something hot, it will recognize that you are trying to pinpoint the location of a specimen and the rate is not coming as background and it will not engage an adjustment action.
I added a whole bunch of features around this concept, such as exponential change of the audio frequency where small change of the rate will cause a progressively large swing in the tone pitch referenced to the squelch level.
Another feature for the tone generation is to start the frequency range at the squelch level so tone's frequency is relative to the squelch level and not absolute to the actual rate - this way if you are working around high level source, tone frequency will still start low and increase with the rate ..and so on.

[Jim Kovalchick]

I recently made an improvement to my rock probe. For my prototype, I used pvc pipe because it is cheap and easy to get, but pvc is not optimum for gamma permeability. I bought some thin wall aluminum tube and now have much improved sensitivity. For low grade uranium rocks, I now see noticeable meter movement several inches further away from the probe.

I also made an addition to my Eberline PRM 6 meter using a Pi Pico microcontroller. I took the signal that normally goes straight to the PRM speaker and set it to input on one of the GPIO points. My code interprets the pulse rate and converts it to a tone whose frequency varies with pulse frequency. This tone is output to the units speaker. I added a LIPO battery and charger shield to the Pico. I repurposed the PRM's speaker on/off switch to toggle the charger shield. When the board first charges up, it reads the base frequency for five seconds and essentially sets a background. In this way, I can use the switch to reset the tone to different background readings as I move about.